Radiation damage in lithium ceramic materials can severely degrade thermal transport properties, limiting their performance in nuclear environments such as tritium-producing burnable absorber rods (TPBARs). This study compares the intrinsic thermal conductivity degradation in single crystals of LiAlO2 and LiAl5O8 due to radiation-induced point defects. LiAlO2 shows a significant drop in thermal conductivity of up to 75% under increasing defect concentration and temperature, while LiAl5O8 retains over 50% of its thermal conductivity, even at high defect levels and elevated temperatures. The greater resilience of LiAl5O8 is attributed to its structural resilience, which suppresses defect generation and preserves phonon transport. Partial phonon density of states analysis reveals that Li and Al vacancies strongly suppress vibrational modes in LiAlO2, while LiAl5O8 shows minimal change, supporting its superior radiation tolerance. These results suggest LiAl5O8 to be a more durable candidate for high-temperature radiation environments.

Abstract Understanding irradiation damage and tritium transport in LiAlO2 ceramics is essential for their deployment in tritium-producing burnable absorber rods (TPBARs). Grain boundaries (GBs) play an important role in governing radiation response and tritium transport in LiAlO2 and its secondary phase LiAl5O8. Using molecular dynamics simulations, we investigated defect evolution and tritium diffusion during displacement cascades in single-crystal and bicrystal LiAlO₂ and LiAl₅O₈ at 600 K. The results reveal that GBs suppress the damage in the LiAlO2 bicrystals as compared to single crystals, via interstitial emission. The effect is more pronounced for Li defects, than for Al and O. In LiAl₅O₈, the damage reduction due to the introduction of GB is less pronounced overall. Tritium diffusion coefficients at GBs are higher by factors of 2–10 relative to diffusion inside the grain bulk. The effect is particularly pronounced in LiAlO₂, where Σ5, Σ17, and Σ25 GBs promote rapid tritium transport, whereas LiAl₅O₈ exhibits slower diffusion due to reduced free volume at its GBs. These findings evince a trade-off that while GBs mitigate radiation damage by absorbing interstitials, they simultaneously provide fast pathways for tritium migration, which is not desirable for TPBAR application. The mechanistic insights gained here establish a foundation for microstructural design strategies to balance radiation tolerance and tritium retention in ceramic breeder materials.

Designing radiation-tolerant high entropy alloys (HEAs) traditionally requires thousands of expensive atomistic simulations with domain knowledge informed iterations. Here, we present an agentic AI framework that autonomously explores alloy compositional space by coupling a large language model (LLM) with on-the-fly molecular dynamics (MD) simulations. The agent operates in a closed-loop manner wherein it proposes new alloy compositions, launches LAMMPS primary knock-on atom simulations on each candidate, extracts defect production, and updates its search strategy based on the evolving defect landscape. Using six distinct starting chemistries, i.e. Co-rich, Cr-rich, Fe-rich, Mn-rich, Ni-rich, and equiatomic, our AI agent adjusts exploration breadth, performing large global sweeps when early simulations indicate high defect production and narrowing to local searches when favorable low defect regions are discovered. The framework demonstrates the ability of an agentic AI framework to autonomously navigate high dimensional compositional space and reason about defect trends without predefined heuristics.

We present an autonomous materials discovery framework that couples a large language model (LLM) with molecular dynamics (MD) simulations to optimize Fe-Cr-Mn alloy compositions for tensile strength. Starting from six distinct compositions, the LLM operated as an intelligent agent, iteratively proposing changes based on prior simulation results and constraints. Over 50 iterations per case, the LLM adaptively explored the composition space, identifying high-strength regions, not easily accessible by conventional methods. The highest strength, 18.7 GPa, was achieved with Fe71Cr25Mn4 composition, identified from a Fe75Cr20Mn5 starting point. The LLM autonomously adjusted its strategy in real time, demonstrating closed-loop decision-making using commodity hardware. This approach showcases the potential of LLMs as scientific co-pilots, capable of accelerating materials discovery and generalizable to other domains like biology and drug design.

Gamma-phase lithium aluminate (gamma-LiAlO2) is used in tritium-producing burnable absorber rods (TPBARs), where efficient thermal transport is essential for thermal stability and tritium release. Under irradiation, gamma-LiAlO2 partially transforms into LiAl5O8, producing a multiphase microstructure whose influence on heat transport remains poorly understood. In this work, molecular dynamics simulations are employed to investigate interfacial thermal resistance in LiAlO2and LiAl5O8. Bicrystal models are used to evaluate Kapitza resistance (Rk) across grain boundaries (GBs) over a range of coincident site lattice misorientations. The results show that Rk increases with misorientation angle before appearing to saturate at higher angles, consistent with enhanced phonon scattering from increased interfacial disorder. The Kapitza resistance ranges from 4.8 & times; 10-11 to 1.1 & times; 10-10 m2 & sdot;K/W for LiAlO2and from 3.9 & times; 10-11 to 6.1 & times; 10-11 m2 & sdot;K/W for LiAl5O8, indicating stronger GB-induced thermal resistance in LiAlO2. Heterophase interfaces between LiAlO2 and LiAl5O8 are constructed to quantify interphase thermal resistance. The interphase Rk spans a range of 0.74 to 1.74 & times; 10-10 m2 & sdot;K/W, substantially higher than GB resistances within the individual phases. These findings demonstrate that irradiation-induced phase transformation and the resulting interphase boundaries play a dominant role in limiting thermal transport in Li-based ceramics.

Lithium aluminate ceramics, LiAlO2 and LiAl5O8, show promise in nuclear environments due to their excellent radiation tolerance. Molecular dynamics simulations investigate grain boundaries (GB) and their role in defect evolution. Results reveal that GBs act as efficient defect sinks, with Li and Al atoms exhibiting distinct behaviors during displacement cascades. Tritium migration in LiAlO2 is also studied, showing rapid diffusion and stable configurations with oxygen, corroborated by ab initio simulations from the literature. The calculated tritium diffusion coefficient of 1.33 × 10−¹⁴ m²/s aligns with the literature, validating the model. LiAl5O8 demonstrates superior defect healing compared to LiAlO2, attributed to enhanced atomic transfer between grains and GBs. These findings reveal key insights into defect dynamics, providing essential insights for their application in tritium-producing burnable absorber rods (TPBARs).

The effect of 120 keV He++D+-ion irradiation on the thermal conductivity of ceramic tetragonal gamma- LiAlO 2 is studied with time-domain thermoreflectance (TDTR) at temperatures between 300 and 700 K. The thermal conductivity of single crystal gamma- LiAlO 2 is 13.5 W/(m & sdot;K) at 300 K, and scales with temperature like 1/T. The thermal conductivity of unirradiated polycrystalline gamma- LiAlO 2 is 7.4 W/(m & sdot;K). Irradiation at fluences of 1 x, 5 x, and 10 x 1016 ions/cm2 decreases the thermal conductivity by approximate to 30 %, 80 %, and 90 %. The effect of irradiation is saturated at ion fluences of 1017 ions/cm2. Irradiation decreases the temperature dependence of the thermal conductivity. For ion fluences larger than 1017 ions/cm2, the thermal conductivity reaches a minimum value of approximate to 1 W/(m & sdot;K) that is independent of temperature.

We have developed a long short-term memory stacked ensemble (LSTM-SE) surrogate modeling approach that can provide rapid predictions of microstructural evolution and the resultant mechanical properties of American Iron and Steel Institute (AISI) 316L series stainless steel (SS316L) fuel cladding under conditions of varying temperature and radiation dose rate. To acquire training data, we developed and implemented a kinetic Monte Carlo (KMC) model to simulate precipitation kinetics of M23C6, gamma', and G phases within SS316L cladding. Experimentally reported precipitation kinetics of SS316L in literature were linked to the kinetic parameters of the simulated precipitation in our KMC model. The model was then used to simulate microstructure evolution under synthetically generated treatments of varying temperature and radiation dose rate for periods of up to 3000 h. Changes in volume fraction, number density, and particle size of precipitates were recorded, and particle area fractions were correlated using statistical methods to develop the surrogate model. Simultaneously, the mechanical properties of the simulated microstructures were evaluated using microstructure-based finite element method (FEM) analysis to determine the elastic modulus, yield stress, ultimate tensile strength, and elongation to failure of the aged microstructures. Using this approach, our surrogate model can predict precipitation behavior within 0.25 % volume fraction and mechanical properties within 6 % relative error from the values predicted by the KMC and FEM models using 50 training simulations as input. The trained recurrent neural network-based model can return estimations of precipitation kinetics and mechanical properties similar to 1000 times faster than the physics-based codes. This work demonstrates, as a proof of concept, that microstructural evolution under variable conditions can be predicted using a statistics-based model informed by a practicably obtainable dataset. The potential applications of this type of modeling framework are discussed.

This review explores molecular dynamics simulations for studying radiation damage in Tritium Producing Burnable Absorber Rod (TPBAR) materials, emphasizing the role of interatomic potentials in displacement cascades. Recent machine learning potentials (MLPs), trained on quantum data, enhance prediction accuracy over traditional models like EAM. We highlight temperature, PKA energy, and composition effects on damage evolution in TPBAR components, recommending suitable potentials and discussing advancements for materials in extreme radiation environments.

This study investigates the impact of Mg and Ni doping on tritium diffusion in LiAlO2 and LiAl5O8 ceramics, that are used in tritium-producing burnable absorber rods (TPBARs). Utilizing Centipede simulations across a broad temperature range (500 K to 1250 K), we explore the interplay between defect dynamics, cluster formation, and tritium mobility. In LiAlO2, Mg doping significantly enhances tritium diffusivity by increasing tritium interstitial concentrations and diffusion coefficients of key species, thereby doubling the overall tritium diffusivity. Ni doping, while shifting the dominant defect to Li vacancies, maintains high tritium mobility due to the low binding energy of Li vacancy-tritium complexes, which ensures effective tritium migration. In LiAl5O8, Mg and Ni doping results in a slight reduction in the diffusion coefficients of key species, yet the dramatic increase in tritium interstitial concentrations compensates, leading to a net small increase in tritium diffusivity. The findings highlight the critical role of defects in tritium transport and the effect of Mg and Ni defects on the performance of these ceramics in demanding nuclear environments.

Carbon fiber composite can be a potential candidate for replacing metal-based battery enclosures of current electric vehicles (E.V.s) owing to its better strength-to-weight ratio and corrosion resistance. However, the strength of carbon fiber-based structures depends on several parameters that should be carefully chosen. In this work, we implemented high throughput finite element analysis (FEA) based thermoforming simulation to virtually manufacture the battery enclosure using different design and processing parameters. Subsequently, we performed virtual crash simulations to mimic a side pole crash to evaluate the crashworthiness of the battery enclosures. This high throughput crash simulation dataset was utilized to build predictive models to understand the crashworthiness of an unknown set. Our machine learning (ML) models showed excellent performance (R2 > 0.97) in predicting the crashworthiness metrics, i.e., crush load efficiency, absorbed energy, intrusion, and maximum deceleration during a crash. We believe that this FEA-ML work framework will be helpful in down select process parameters for carbon fiber-based component design and can be transferrable to other manufacturing technologies.

This paper presents a probabilistic surrogate model for the accelerated design of electric vehicle battery enclosures with a focus on crash performance. The study integrates high-throughput finite element simulations and Gaussian Process Regression to develop a surrogate model that predicts crash parameters with high accuracy while providing uncertainty estimates. The model was trained using data generated from thermoforming and crash simulations over a range of material and process parameters. Validation against new simulation data demonstrated the model's predictive accuracy with mean absolute percentage errors within 8.08 Carlo uncertainty propagation study revealed the impact of input variability on outputs. The results highlight the efficacy of the Gaussian Process Regression model in capturing complex relationships within the dataset, offering a robust and efficient tool for the design optimization of composite battery enclosures.

Austenitic 347H stainless steel offers superior mechanical properties and corrosion resistance required for extreme operating conditions such as high temperature. The change in microstructure due to composition and process variations is expected to impact material properties. Identifying microstructural features such as grain boundaries thus becomes an important task in the process-microstructure-properties loop. Applying convolutional neural network (CNN) based deep-learning models is a powerful technique to detect features from material micrographs in an automated manner. Manual labeling of the images for the segmentation task poses a major bottleneck for generating training data and labels in a reliable and reproducible way within a reasonable timeframe. In this study, we attempt to overcome such limitations by utilizing multi-modal microscopy to generate labels directly instead of manual labeling. We combine scanning electron microscopy (SEM) images of 347H stainless steel as training data and electron backscatter diffraction (EBSD) micrographs as pixel-wise labels for grain boundary detection as a semantic segmentation task. We demonstrate that despite producing instrumentation drift during data collection between two modes of microscopy, this method performs comparably to similar segmentation tasks that used manual labeling. Additionally, we find that na\"ive pixel-wise segmentation results in small gaps and missing boundaries in the predicted grain boundary map. By incorporating topological information during model training, the connectivity of the grain boundary network and segmentation performance is improved. Finally, our approach is validated by accurate computation on downstream tasks of predicting the underlying grain morphology distributions which are the ultimate quantities of interest for microstructural characterization.

Molecular dynamics was employed to investigate the radiation damage due to collision cascades in LiAlO2 and LiAl5O8, the latter being a secondary phase formed in the former during irradiation. Atomic displacement cascades were simulated by initiating primary knock-on atoms (PKA) with energy values = 5, 10 and 15 keV and the damage was quantified by the number of Frenkel pairs formed for each species: Li, Al and O. The primary challenges of modeling an ionic system with and without a core–shell model for oxygen atoms were addressed and new findings on the radiation resistance of these ceramics are presented. The working of a variable timestep function and the kinetics in the background of the simulations have been elaborated to highlight the novelty of the simulation approach. More importantly, the key results indicated that LiAlO2 experiences much more radiation damage than LiAl5O8, where the number of Li Frenkel pairs in LiAlO2 was 3–5 times higher than in LiAl5O8 while the number of Frenkel pairs for Al and O in LiAlO2 are ~ 2 times higher than in LiAl5O8. The primary reason is high displacement threshold energies (Ed) in LiAl5O8 for Li cations. The greater Ed for Li imparts higher resistance to damage during the collision cascade and thus inhibits amorphization in LiAl5O8. The presented results suggest that LiAl5O8 is likely to maintain structural integrity better than LiAlO2 in the irradiation conditions studied in this work.

Molecular dynamics was employed to investigate the radiation damage due to collision cascades in LiAlO 2 and LiAl 5 O 8 , the latter being a secondary phase formed in the former during irradiation. Atomic displacement cascades were simulated by initiating primary knock-on atoms (PKA) with energy values = 5, 10 and 15 keV and the damage was quantified by the number of Frenkel pairs formed for each species: Li, Al and O. The primary challenges of modeling an ionic system with and without a core–shell model for oxygen atoms were addressed and new findings on the radiation resistance of these ceramics are presented. The working of a variable timestep function and the kinetics in the background of the simulations have been elaborated to highlight the novelty of the simulation approach. More importantly, the key results indicated that LiAlO 2 experiences much more radiation damage than LiAl 5 O 8 , where the number of Li Frenkel pairs in LiAlO 2 was 3–5 times higher than in LiAl 5 O 8 while the number of Frenkel pairs for Al and O in LiAlO 2 are ~ 2 times higher than in LiAl 5 O 8 . The primary reason is high displacement threshold energies (E d ) in LiAl 5 O 8 for Li cations. The greater E d for Li imparts higher resistance to damage during the collision cascade and thus inhibits amorphization in LiAl 5 O 8 . The presented results suggest that LiAl 5 O 8 is likely to maintain structural integrity better than LiAlO 2 in the irradiation conditions studied in this work.

Tritium (T) and He diffusion in LiAlO2 and LiAl5O8 phases influences the performance of tritium producing burnable absorber rods (TPBARs) by affecting the gas release, swelling and thermal conductivity of Li-bearing ceramic pellets. Frenkel pair defects and clusters created by irradiation can attract T and He interstitials and form clusters of the type HeixLi,HeixAl,HeixO,TixLi,TixAlandTixO,1≤x≤4 in a Li, Al or O vacancy site (notation denotes x He or T atoms in a 1 Li, 1 Al or 1 O vacant site). The concentration and mobility of each of these clusters collectively contribute to the diffusion of the He and T gases in LiAlO2 and LiAl5O8. In this work, free energy cluster dynamics simulations implemented in the Centipede code, are used to obtain the concentration and diffusivities of these clusters which are then used to calculate the total diffusivity of T and He gases in LiAlO2 and LiAl5O8. The results show that diffusivity of T is at least one order of magnitude higher in LiAlO2 as compared to that in LiAl5O8 whereas He diffusion is 2–13 orders of magnitude higher in LiAlO2 as compared to that in LiAl5O8. There is a higher concentration of highly diffusive species (T interstitials and Ti03Li for the case of tritium and Hei01Li, Hei02Li and Hei03Li for the case of He) in LiAlO2 than in LiAl5O8 which increase the total diffusion of T and He in LiAlO2.

Background: Multi-Principal Element Alloys (MPEAs) have better properties, such as yield strength, hardness, and corrosion resistance compared to conventional alloys. Compositional optimization is a challenging task to obtain desired properties of MPEAs and machine learning is a potential tool to rapidly accelerate the search and design of new materials. Methods: We have implemented different machine learning models to predict the yield strength and Vickers hardness of MPEAs at room temperature and quantify the uncertainty of the predictions. Results: Our results suggest that valence electron concentration (VEC) is the key feature dominating the yield strength and hardness of MPEAs. Our predicted yield strength and hardness values for the experimental validation set show < 15 % error for most cases with respect to the experimental values. Conclusions: Our machine learning model can serve as a useful tool to screen half a trillion MPEAs and down select promising compositions for useful applications.

A Monte Carlo simulation method capable of replicating the kinetics of M23C6 precipitation in 347H stainless steels was developed for the purpose of producing synthetic microstructures that approximate its microstructural evolution under aging periods of up to 10,000 hours at temperatures between 600 °C and 750 °C. To accomplish this, experimental data from the literature was used to parameterize simulations and replicate the nucleation and growth kinetics of M23C6 particles within 347H and similar austenitic stainless steel alloys. These simulations were found to have considerable fidelity to previous efforts to study the precipitation of M23C6 in other 300 series stainless steel alloys. Synthetic 347H microstructures were then generated that accounted the effects of aging temperature, duration, dislocation density, and the presence of boron within the microstructure. These simulations predict several key trends, those being that (1) the size of M23C6 precipitates decreased with aging temperature and (2) the growth rate of M23C6 particles decreased with aging temperature. Further, while (3) the addition of dislocation density due to creep conditions resulted in increasing intragranular nucleation of M23C6 precipitates with increasing dislocation density and (4) B additions within the microstructure led to modest increases in precipitate size above 700 °C, which indicates that more complex physics are necessary to account for the presence of B.

With a goal of exploiting additive manufacturing to improve the manufacturing of existing reactor materials, we developed a chemical composition-based machine learning model to predict the printability of any given alloy in laser powder bed fusion (L-PBF) using experimental data from peer-reviewed literature. We defined printability as the ability to avoid defects like cracking, balling, porosity, and lack of fusion, that are caused by thermal stresses (during solidification or liquation), molten pool disintegration into disconnected small beads or lack of heat input respectively. Our models predict the tendency of balling defect formation and porosity percentage for a given composition, under a given set of processing conditions. To predict the likelihood of balling defect, three models: a random forest classifier, a gradient boost regressor and a neural network were trained on a dataset containing both single and multi principal element alloys. The neural network model showed the highest accuracy of 92.3 % in predicting the balling defect formation. A random forest regressor, gradient boost regressor and neural network were trained and tested on a dataset of various alloys to predict porosity. The random forest regressor showed the best predictions with an R2 score of 0.97. The models also revealed the relative importance of the input descriptors on defect-formation tendency. Of particular significance was the identification of carbon as an important element in determining the occurrence of balling and percent porosity in alloys like steel, as well as being moderately important to the percentage porosity in other alloys as well as steel. Manganese was also identified as a key descriptor for the percentage of porosity in steel and other alloys. Manganese's low thermal conductivity and consistent presence in the dataset is the likely cause for its contribution. Carbon's role is attributable to its relatively high specific heat and high melting temperature. Our model serves as a swift, chemistry-based tool to design experiments and find modified compositions better suited for additive manufacturing.

This work investigates the crashworthiness of carbon fiber ply-based electric vehicle’s (EV) battery enclosure, which is a large component currently made using aluminum alloys. A finite element analysis based framework was used to perform the thermoforming simulation followed by as-formed structural analysis to examine the strength of carbon fiber plies in the structural design of an EV battery enclosure. Simulations were performed for the thermoforming process of the enclosure panel and side pole impact test of the as-formed part. In step one, using a thermoforming simulation, carbon fiber layers were formed on the surface of a die representing the geometry of the desired battery enclosure. In step two, a side pole impact test was simulated to investigate the crashworthiness of the battery enclosure for various impact speeds. Different crashworthiness characteristic parameters such as peak load, average load, crush load efficiency, energy absorbed, material damage, and material deformation were calculated from the results. The effect of impact speed and the number of carbon fiber layers on all the crashworthiness characteristic parameters were studied in detail. The behavior of the enclosure obtained by the crash test was compared with the behavior of an enclosure made from traditional aluminum alloys and the crashworthiness of both materials was found to be similar for a wide range of impact speeds. An isotropic elasto-plastic material model along with FLD (Forming Limit Diagram) damage model was used for aluminum alloy material.